What was billed, what was authorized, what was actually delivered — and what the DRG was paid on.
Utilization and payment integrity are the same question asked at two moments: before the service, whether it should happen; after the claim, whether what was paid matches what happened — and whether the DRG it was paid on is supported. All of it lives in data you already hold, and most of it is rented, often on contingency.
The prior-authorization metrics the rule now requires
Federal rules put decision timeframes and public metric reporting on a fixed clock. Alberto produces the pack: volume and outcomes by urgency, request type, procedure and care setting, plus turnaround against the standard — computed, not estimated.
Where a required metric depends on data your systems don’t yet capture, the artifact says so explicitly and names the gap. A reported number with nothing behind it is worse than an acknowledged blank.
The authorizations that aren’t doing any work
Prior authorization is a cost lever in both directions: it manages utilization, and it consumes staff time and physician goodwill. Alberto identifies the procedure and setting combinations where authorization almost never changes the outcome — a quantified removal list, which is precisely the burden reduction the rule intends.
In practice, approval rates vary enormously by setting. Authorization does real work in a few of them. The rest is friction you can price.
Readmissions on the federal definition
Unplanned readmissions computed on the published methodology, for every member enrolled at admission rather than only those still active — an active-member denominator silently drops a large share of qualifying stays. Observed rates only, with a discharge month reported only once every window inside it has closed.
DRG validation — what the paid tier actually rests on
An inpatient claim is paid on a DRG, and the DRG’s tier is often set by a complication or comorbidity. Whether that complication was present on admission or acquired in the hospital is the whole question — and it is answered by a flag the hospital itself supplies.
Alberto reads every paid inpatient stay for the secondary diagnoses that drove the tier and separates three cases cleanly: the ones federal payment rules already address, the ones that depend on your contract terms, and the ones that are a documentation-completeness question and never a finding. Per hospital, per DRG family, with the paid tier, the supported tier and the difference.
Financial impact is labeled an estimate, because that is what it is.
Fraud, waste and abuse — with the control run first
Before any anomaly list is produced, the platform runs the control that tests whether the pattern is a provider behavior or an artifact of our own data. Lists that skip that step end with an accusation against a contracted provider and a retraction. That order is not optional here.
Live today: transport and ancillary anomaly review — the trip that couldn’t have happened, the mileage that doesn’t match the map, the leg billed twice — with the geography checked before anyone is named. And the DRG work above, which is where a great deal of hospital waste actually lives.
Everything a contingency vendor would find in this data, it should be finding for you, at a flat cost.
Computed, not estimated — and honest about what your systems don’t capture yet.